- Email: [email protected]
Immunoblot analyses of glycogen debranching enzyme in different subtypes of glycogen storage disease type III
Immunoblot analyses of glycogen debranching enzyme in different subtypes of glycogen storage disease type III Jia-Huan Ding, MD, PhD, T. d e Barsy, MD, Barbara I. Brown, PhD, Rosalind A. C o l e m a n , MD, a n d Yuan-Tsong Chen, MD, PhD From the Division of Pediatric Genetics and Metabolism, Duke UniversitY Medical Center, Durham, North Carolina, the International Institute of Cellular and Molecular Pathology, Universit~ Catholique de Louvain, Brussels,Belgium, and the Department of Biological Chemistry and Molecular Biophysics, Washington University School Of Medicine, St, Louis, Missouri TO determine the tissue distributionof glycogen debranching enzyme, we used immunoblot analYSiS witha polyclonal antibody prepared against purified porcine muscle debranching enzyme. Debranching enzyme was identified in porcine brain, kidney, (:ardiac muscle, skeletal muscle, liver, and spleen; and in human liver, skeletal muscle, lymphocytes, lymphoblastoid cells, skin fibroblasts; cultured chorionic villi, and amniocytes. !n each of these tissues the debranching enzyme band was 160 kd. To determine the molecular basis for glycogen storage disease type III at the protein level, tissues from 41 patients with glycogen storage disease type III were also subjected to immunoblot analysis. Three patients having isolated transferase deficiency with retention of glucosidase activity (type IIID disease) had nearly normal amounts of cross-reactive material. In the remaining patients (both transferase and glucosidase deficiency), debranching enzyme was either absent or greatly reduced. These latter patients included 31 with disease that appeared to involve both liver and muscle (type IliA), four with disease that was present only in the liver (type IIIB), and three with unknown muscle status. In patients with both type IliA and type IIIB disease, debranching enzyme protein was absent in skin fibroblasts, lymphoblastoid cells, and lymphocytes. The parents of two patients with type IliA disease had art intermediate level of debranching enzyme protein, consistent with their presumed heterozYgote State. An immunoblot analysis of cultured amniotic fluid cells from a woman whose fetus was at risk for type IliA disease predicted an unaffected fetus; the prediction was confirmed postnatally. Thus Western blot analysis offers an alternate method of prenatal diagnosis for the most common form of glycogen storage disease type III. (J PEDIATR1990;116:95-
100)
Supported by National Institutes of Health grants DK39078 (Dr. Chen), DK09235 (Dr, de Barsy), and M01-RR30, Division of Research Resources General Clinical Research Centers Program, a grant from the Muscular Dystrophy Association (Dr. Chen), and a generous contribution to the Duke GSD fund by E. B..Mandel. Submitted for publication June 1, 1989; accepted.July 1, 1989. Reprint requests: Yuan-Tsong Chen, MD, PhD, Duke University Medical Center, Department of Pediatrics, Box 3028, Durham, NC 27710. 9/20/15459
Glycogen debranching enzyme is a eukaryotic enzyme that contains two catalytic activities on a Single polypeptide chain. The two Catalytic activities, olig0-1,4-1,4-glucantransferase (EC 2.4.1.25) an d amylo-l,6-gluc0sidase (EC 3.2.1.33), occur at separate catalytic sites on the polypeptide chain and can function independently of each other.a, 2 Deficiency of glycogen debranching enzyme activity causes glycogen storage disease type III (Cori disease, limit dextrinosis). Patients with type III disease vary remarkably, 95
96
Ding et al.
CRM
The Journal of Pediatrics January 1990
Cross-reactivematerial
I
both clinically and enzymatically.3"5 Som~ patients have li~,er involvement manifested by hepatomegaly and hypoglycemia; a very few patients have muscle weakness and wasting but no clinically apparent liver disease; and many patients have problems related to both liver and muscle. Van Hoof and Hers 3 measured the combined and separate activities of transferase and glucosidase in both liver and muscle, and divided type III disease into six biochemical subgroups, s In this classification, most patients have had very low or absent debranching enzyme activity in both liver and muscle (type I l i A disease), and some have had enzyme deficiency limited to the liver (type IIIB disease). Some patients have had an intermediate level of enzyme activity, and in rare instances a selective loss of only one of the two debranching activities (glucosidase or transferase) has been demonstrated. In a study of 34 patients, Brown and Brown 4 found disease subtypes I l i A and IIIB and one example of possible isolated glucosidase deficiency, but they were unable to identify cases with intermediate levels of enzyme activity. It appears that 75% to 85% of patients with glycogen storage disease type III have generalized deficiency, affecting both liver and muscle (type I l i A disease), and 17% have the enzyme activity absent in the liver but retained in the muscle (type IIIB disease); only rarely do patients have a selective loss of one of the two activities.46 The molecular basis for the clinical and enzymatic variability in glycogen storage disease type III is not understood. We therefore initiated studies of debranching enzyme to elucidate the molecular mechanism and to identify the basis for the subtypes observed enzyrnatically. We previously reported the purification of glycogen debranching enzyme, characterization of antibody, and absence of cross-reactive material in five patients with type I l i a disease. 7 In this study, we extended our investigation of the debranching enzyme to a total of 41 patients with various subtypes, and we now report on the enzyme from a variety of tissues. We also report the first use of immunoblot analysis for the detection of heterozygotes and for prenatal diagnosis. METHODS Reagents. Nitrocellulose paper was purchased from Schleicher & Schuell Inc. (Keene, N.H.); iodine 125labeled protein A was obtained from Amersham Corp. . (Arlington Heights, t11.); and Coomassie brilliant blue, protein-molecular-weight standards, and reagents for polyacrylamide gels were from Bio-Rad Laboratories (Richmond, Calif.). Tissue culture supplies and fetal calf serum were from Gibco Laboratories (Grand Island, N.Y.). Phenylmethyisulfonyl fluoride, ethylenediaminetetraacetate,
and ethyleneglycoltetraacetate were from Sigma Chemical Co. (St. Louis, Mo.). Patients. We studied 41 patients with glycogen storage disease type III whose ages ranged from 4 months to 50 years. Type III disease had been previously diagnosed by demonstrating deficient debranching enzyme activity in liver or muscle biopsy samples, or both, with phosphorylase limit dextrin used as the substrate. 8 This method measures combined oligo-l,4-1,4-glucantransferase and amylo-l,6glucosidase activities. Incorporation of carbon 14-labeled glucose into glycogen was also measured in 15 patients. 9 In 24 cases, debranching enzyme activity was measured in both liver and muscle. By the Van Hoof-Hers enzymatic classification,3 19 patients had disease that involved both liver and muscle (type I l i A ) , four had disease confined to the liver (type IIIB), and three had isolated transferase deficiency in liver and muscle (type IIID disease). Type IIID disease is defined as deficient debranching enzyme activity with phosphorylase limit dextrin but normal or slightly reduced incorporation of [14C] glucose into glycogen. In the remaining 15 patients, debranching enzyme activity was measured only in liver; in 12 of these patients, the clinical findings suggested that muscle involvement was also present; in the other three patients, the clinical data were incomplete. Tissue samples. All human liver and muscle biopsy samples used in our investigation had been stored at - 2 0 ~ C or - 7 0 ~ C (some for >2 years) after the original enzymatic or pathologic diagnosis was made. Control samples were from patients with other metabolic diseases (cystic fibrosis; glycogen storage disease types I, II, VI, and IX; unknown myopathy). Chorionic villi and amniotic cells were obtained from pregnant women whose fetuses were at risk for chromosomal abnormalities. Fresh porcine tissues were obtained from a local abattoir. Cell lines. Four skin fibroblast cultures (GM 02523, G M 03390, GM 00576, GM 00573) from patients with glycogen storage disease type III were obtained from the Institute of Medical Research, Camden, N.J. Six skin fibroblast cultures and eight lymphoblast0id cell lines from patients with type III disease were established in our laboratory after informed consent for skin biopsy and blood drawing was obtained. Lymphocytes were separated from peripheral blood samples and transformed by Epstein-Barr virus to produce permanent cell lines. The fibroblast cultures were maintained in Eagle minimum essential medium with 10% fetal calf serum, and the lymphoblastoid cells were grown in RPMI 1640 culture medium (Gibco) with 10% fetal calf serum. Penicillin, 100 units/ml, and streptomycin, i 00 #g/ ml, were added to the culture media.
Immunoblot analysis Antiserum. The antiserum against purified porcinemuscle debranching enzyme was produced in New Zealand
Volume 116 Number I
G l y c o g e n d e b r a n c h i n g e n z y m e in glycogen s t o r a g e disease
rabbits. This antiserum reacts against porcine debranching enzyme with a single precipitation line and inhibits the debranching enzyme activity] Because our antiserum recognized nonspecific bands in fibroblasts, lymphoblastoid cells, and lymphocytes, the antiserum was affinity-purified with purified porcine enzyme before immunoblot analyses of these cells. Antiserum (0,5 ml) was added to a nitrocellulose disk (1.5 cm in diameter) containing 20 ~zg of purified debranching enzyme and incubated for 1 hour. The filter was washed with phosphate-buffered saline solution, and antibody was eluted with 0.5 sodium chloride, 0.2 mol/L acetic acid, and then neutralized. W e s t e r n blots. Tissue samples were homogenized and cultured cells were sonicated in the presence of 5 volumes of buffer containing 4 mmol/L ethylenediaminetetraacetare, 2 mmol/L ethyleneglycoltetraacetate (pH 7.2), and 0.34 mmol/L phenylmethylsulfonyl fluoride. The homogenate or sonicate was then centrifuged at 10,000g for 15 minutes, an d the supernatant was subjected to immunoblot analysis. Up to 80 ~zg of protein from the supernatant was applied to 7% sodium dodecylsulfate-potyacrylamide gel electrophoresis according to the method of Laemmli) ~ The electrotransfer of protein and the immunologic detection of protein were performed as described elsewhere, 11 with the following modifications: The proteins were electroblotted onto nitocellulose paper by means of transblotter equipment (Bio-Rad Laboratories). The nitrocellulose filter was washed with 5% milk (Carnation nonfat dry milk) in phosphate-buffered saline solution for 1 hour. The antiserum or antibody (1:100 dilution in fresh 5% milk) was then added, and the solution was incubated for 2 hours. The filter was washed for 1 hour with three changes of 1% milk and incubated with 5 X l0 s cpm/ml [1251] protein A for 2 hours. The filter was then washed three times with 1% milk and once with 0.5% Triton X-100 detergent, each time for 15 minutes. The filter was then dried and autoradiographed. Protein was measured by the Bradford method, 12 with bovine serum albumin used as the standard. RESULTS Tissue distribution of debranching enzyme. Immunoblot analysis with anti-porcine-muscle debrancher antiserum showed that this antiserum can detect debranching enzyme in porcine kidney, heart, skeletal muscle, liver, brain, and spleen (data not shown). The major band detected in each of these tissues was 160 kd, the same size as the purified debranching enzyme. The amount of debfancher protein was highest in liver, heart, and skeletal muscle, and low in spleen, kidney, and brain. The antibody als0 detected debrancher protein in cultured chorionic villi cells, lymphocytes, lymphoblastoid cells, skin fibroblasts, and cultured amniotic fluid cells (Fig. 1). The amount of debrancher protein in these tissues was small in comparison with that
]
2
3
4
5
6
7
97
8
tOO
16 -93
-66
,~
-45
Fig. t. Distribution of debranching enzyme m human tissues obtained by immunoblot analysis of protein from various human tissues probed with affinity-purified antibody against muscle debranching enzyme, Lane I, Erythrocytes; lane 2, cultured chorionic villi ceils; lane 3, lymphoeytes; lane 4, lymphoblastoid cells; lane 5, muscle tissue; lane 6, cultured fibroblasts; lane 7, cultured amniotic cells; and lane 8, liver tissue. In lanes I to 4, 6, and 7, each well contained 80 gg of protein. Lanes 5 (muscle) and 8 (liver) each contained 20/zg of protein. Numbers on right are molecular masses, measured in kilodaltons, of marker proteins.
in liver or muscle. As estimated from densitometry scanning and the amount of protein applied to the immunoblot, the amount of debrancher protein in these tissues was less than 10% of the amount that was present in muscle or liver. Debrancher protein in patients with glycogen storage disease type IIL Forty-eight samples from 41 patients with glycogen storage disease type III were subjected to immunoblot analysis. The analyses were performed on coded samples without prior knowledge of the diagnoses. The samples were from 16 livers, 10 muscles, 3 lymphocyte preparations, 10 skin fibroblast cultures, and 8 lymphoblastold cell lines. In 7 patients, more than one tissue was analyzed: liver and muscle, 4 patients (3 with type I I I A disease and 1 with type IIID disease); fibroblasts and lymphoblastoid cells, 2 patients (type IIIB disease); lymphocytes and fibroblasts, 1 patient (type IIIB disease). Representative samples are shown in Figs, 2 and 3. Deficient debranching enzyme was demonstrated, not only in liver, muscle, or
98
Ding et al.
The Journal o f Pediatrics January 1990
MUSCLE I
2
3
4
LIVER 5
6
7
8
LIVER 9
I0
It
MUSCLE
12 15 14 15 16 17
%:
..... ,,
Fig. 2. Immunoblot analysis of debrancher protein from liver and muscle tissue probed with antiserum against porcine debranching enzyme. Each well contained 40 ~g of protein. Tissues contained in lanes 2, 3, 4, 6, 7, 13, 14, and 17 were from patients with glycogen storage disease type II[. Lane 6 contains tissue from patient with isolated transferase deficiency (type IIID disease), lane 14 from patient with type IIIB disease, and remainder from patient with type IIIA disease. Lane 7 (liver) and lane 17 (muscle) contain tissues from same patient with type IIIA disease. Remaining lanes contain control samples from patients with other metabolic diseases. Arrow indicates position of debranching enzyme.
F I BRQBLASTS 2
~
a
5
6
7
8
9
LYMPHOCYTES t0
~
12
15
~4
15
16
17
IB
19
20
Fig. 3. Immunoblot analysis of debrancher protein from skin fibroblasts and lymphocytes probed with affinity-purified antibody against muscle debranching enzyme. Each well contained 40 #g protein from skin fibroblasts ,lanes 1 to 6 and 8 to 14) or lymphocytes (lanes 15 to 20). Lane 7 contained 10 #g of protein from human muscle as control. Lanes 1, 4, 5, 6, 12, 14, and 18 contained tissue from patients with type IIIA disease, and lanes 9 and 15 from type IIIB disease. Lanes 10 and 11 and lanes 19 and 20 were from parents of patients 12 and 18, respectively. Lanes 2, 3, 8, 13, 16, and 17 are control samples from patients with other metabolic diseases.
both, but also in skin fibroblasts, lymphocytes, and lymphoblastoid cells. In three patients who had nearly normal amounts of CRM, isolated transferase deficiency (type IIID) was diagnosed enzymatically. The patient with type IIID disease, whose findings are shown in Fig. 2, also had C R M present in a muscle sample. In four patients with type IIIB disease, C R M was absent in all four skin fibroblast cultures, in both of the lymphoblastoid cell lines, and in the lymphocyte culture studied. No liver or muscle samples from the four patients with type IIIB disease were available.
The patients with type I I I A disease included 19 with an enzymatic diagnosis made on the basis of both liver and muscle samples and 12 whose diagnosis was made only on the basis of liver samples but who had clinical findings compatible with muscle involvement. In these patients, immunoblot analysis was performed on skin fibroblasts, lymphoblastoid cells, and lymphocytes in 18, and on liver or muscle samples, or both, in the remaining 13 patients. Regardless of the tissue source studied, all patients with type I I I A disease had either absent or a greatly reduced amount
Volume 116 Number l
Glycogen debranching enzyme in glycogen storage disease
of CRM representing debrancher protein. In two patients who had residual CRM in muscle, the debranching enzyme band was no more than 5% of the control by densitometry scanning. We did not detect debranching enzyme of an abnormal size in any patient. For an additional three patients, insufficient data were available to determine muscle status; enzyme activity was deficient in liver, and CRM was absent.
Heterozygote detection and prenatal diagnosis of glycogen storage disease type IIl. Skin fibroblasts and lymphocytes from two families with type I I I A disease were subjected to immunoblot analysis. No debranching enzyme band was detected in the probands (Fig. 3). The parents of the patients had an intermediate level of debranching enzyme protein in cell extracts, consistent with their presumed hcterozygote state (Fig. 3). Prenatal diagnosis was performed in one of the families. In early childhood the proband had had hypoglycemia, hepatomegaly, and failure to thrive. Cultured amniocytes from the pregnancy at risk demonstrated a debranching enzyme band of the same size and intensity as a control amniocyte culture, thus predicting an unaffected fetus (data not shown). This result was confirmed postnatally by immunoblot analysis of the lymphoblastoid cells established after birth. The baby at 1 year of age had no clinical or laboratory evidence of glycogen storage disease type III. DISCUSSION Most patients with glycogen storage disease type III manifest the disease in several organ systems. Symptoms of myopathy and of liver involvement are most frequent, but neural involvement, renal tubular acidosis, splenomegaly, and cardiac enlargement have been described. Although glycogen with short outer branches accumulates in erythrocytes of patients with type III disease, the absence of debranching enzyme CRM in normal erythrocytes suggests that the amount of debranching enzyme in these cells is lower than our method can detect. The presence of debranching enzyme in a variety of tissues is consistent with multiorgan system involvement in type III disease. Regardless of whether enzyme deficiency or clinical disease was identified in both liver and muscle (type I I I A disease) or was confined to liver alone (type IIIB disease), our data demonstrated the skin fibroblasts, lymphocytes, and lymphoblastoid cells had absent CRM. This finding suggests that the regulation of dehranching enzyme expression is similar in liver, fibroblasts, lymphocytes, and lymphoblastoid cells but is different in muscle. The molecular mechanism for the control of tissue-specific expression of debranchifig enzyme is unknown. Regulation is unlikely to be due to the presence of multiple genes that encode isoenzymes, because previous studies of the purified protein provide no evidence for subunits or isoenzymes of debranching enzyme, and because
99
our data show that the debranching enzyme is 160 kd in all tissues examined. Isolated transferase deficiency (type IIID disease) was diagnosed by measuring deficient debranching enzyme activity with phosphorylase limit dextrin but normal or only slightly reduced incorporation of [14C]glucose into glycogen. Although not all patients studied had enzyme activity measured by both methods, we identified three patients who had nearly normal amounts o] CRM, all of whom had isolated transferase deficiency (type IIID disease). The presence of CRM indicates that the principal antigenic determinant regions of the protein are not located near the catalytic center responsible for transferase activity. Thus our demonstration that patients with type IIID disease have CRM of 160 kd strongly supports previous studies which indicated that debranching enzyme contains two catalytic activities at separate sites on a single polypeptide chain. I, 2 At the molecular level, the defects causing absent CRM may be heterogeneous. The mutation may affect the promotor region or splicing unction. The mutation may create a nucleotide substitution that results in an incomplete protein or an unstable protein. An additional possibility is that the absence of CRM may result from a major alteration of the debranching enzyme gene. In any case, the fact that C R M is absent in most patients wtih glycogen storage disease type II1 has a practical implication. Because enzyme activity is relatively low and difficult to measure in readily accessible tissues such as skin fibroblasts, cultured amniotic fluid cells, or leukocytes,6, 13, 14 Western blot analysis offers an alternative method for prenatal and postnatal diagnosis of type III disease, provided that the proband has no CRM for debranching enzyme. If the CRM status of the proband is not known, a combination immunobtot and biochemical analysis 14-16would be most helpful in providing the diagnosis. Although type III disease can be diagnosed in readily accessible tissues, a positive test result will not indicate whether the enzyme deficiency is generalized or confined to the liver. For the initial diagnosis, we recommend that fibroblasts be sampled. At the present time we do not know whether a normal serum creatine kinase level excludes muscle involvement. Therefore the differentiation between type I I I A and IIIB disease will also require analysis of material obtained by muscle biopsy. We are grateful to the Genetic Division of New York University Medical Center for tissue contribution.
REFERENCES 1. Bates EJ, Heaton GM, Taylor C, Kernohan JC, Cohen P. Debranching enzyme from rabbit skeletal muscle: evidence for the location of two active centres on a single polypeptide chain. FEBS Lett 1975;58:181-5. 2. Gillard BK, Nelson TE. Amylo-l,6-glucosidase/4-a-glucan-
10 0
3. 4.
5.
6.
7.
8.
9.
Ding et al.
The Journal of Pediatrics January 1990
otransferase: use of reversible substrate model inhibitors to study the binding and active sites of rabbit muscle debranehing enzyme. Biochemistry 1977;16:3978-87. Van Hoof F, Hers HG. The subgroups of type llI glycogenosis. Biochemistry 1967;2:265-70. Brown BI, Brown DH. Glycogen storage diseases: types I, III, IV, V, VII and unclassified glycogenoses. In: Dickens F, Randie P J, Whelan W J, eds. Carbohydrate metabolism and its disorders; vol 2. New York: Academic Press, 1968;123-60. Howell RR, Williams JC. The glycogen storage diseases. In: Stanbury JB, Wyngaarden JB, Fredrickson DS, Goldstein JL, Brown MS, eds. The metabolic basis of inherited disease. 5th ed. New York: McGraw-Hill, 1983;141-66. Brown BI, Brown DH. Definitive assays for glycogen debranching enzyme in human fibroblasts. In: Schotland DL, ed. Disorders of the motor unit. New York: Wiley, 1982;667-73. Chen YT, He JK, Ding JH, Brown BI. Glycogen debranching enzyme: purification, antibody characterization, and immunoblot analyses of type III glycogen storage disease. Am J Hum Genet 1987;41:1002-15. Brown DH, Brown BI. Enzymes of glycogen debranching: amylo-l,6-glucosidase (I) and oligo-l,4-glucantransferase (II). Methods Enzymol 1966;8:515-24. Hers HG, Verhue W, Van Hoof F. The determination of amylo-l,6-glucosidase. Eur J Biochem 1967;2:257-64.
10. Laemmli UK. Cleavage ofstructural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227: 680-5. 11. Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedures and' some applications. Proc Natl Acad Sci USA 1979;76:4350-4. 12. Bradford M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Chem 1976;72:248-54. 13. Gutman A, Barash V, Schramm H, et al. Incorporation of [14C]glucose into ~-1,4 bonds of glycogen by leukocytes and fibroblasts of patients with type III glycogen storage disease. Pediatr Res 1985;19:28-32. 14. Van Diggelen OP, Janse HC, Smit GPA. Debranching enzyme in fibroblasts, amniotic fluid cells and chorionic villi: pre- and postnatal diagnosis of glycogenosis type III. Clin Chim Acta 1985;149:129-34. 15. Brown BI, Brown DH. Definitive assays for glycogen debranching enzyme in human fibroblasts. In: Schotland DL, ed. Disorders of the motor unit. New York: Wiley, 1982:667-73. 16. Maire I, Mathieu M. Possible prenatal diagnosis of type III glycogenosis. J Inher Metab Dis 1986;9:89-91.
FELLOWSHIPS Available fellowships in pediatric subspecialties and those for general academic pediatric training are listed once a year, in May, in THE JOURNAL OF PEDIATRICS. Each October, forms for listing such fellowships are sent to the Chairman of the Department of Pediatrics at most major hospitals in the United States and Canada. Should you desire to list fellowships, a separate application must be made each year for each position. All applications must be returned to The C.V. Mosby Company by February 15 of the listing year to ensure publication. Additional forms will be supplied on request from the Journal Editing Department, The C.V. Mosby Company, 11830 Westline Industrial Drive, St. Louis, M O 63146-3318/314-872-8370.
100)
Supported by National Institutes of Health grants DK39078 (Dr. Chen), DK09235 (Dr, de Barsy), and M01-RR30, Division of Research Resources General Clinical Research Centers Program, a grant from the Muscular Dystrophy Association (Dr. Chen), and a generous contribution to the Duke GSD fund by E. B..Mandel. Submitted for publication June 1, 1989; accepted.July 1, 1989. Reprint requests: Yuan-Tsong Chen, MD, PhD, Duke University Medical Center, Department of Pediatrics, Box 3028, Durham, NC 27710. 9/20/15459
Glycogen debranching enzyme is a eukaryotic enzyme that contains two catalytic activities on a Single polypeptide chain. The two Catalytic activities, olig0-1,4-1,4-glucantransferase (EC 2.4.1.25) an d amylo-l,6-gluc0sidase (EC 3.2.1.33), occur at separate catalytic sites on the polypeptide chain and can function independently of each other.a, 2 Deficiency of glycogen debranching enzyme activity causes glycogen storage disease type III (Cori disease, limit dextrinosis). Patients with type III disease vary remarkably, 95
96
Ding et al.
CRM
The Journal of Pediatrics January 1990
Cross-reactivematerial
I
both clinically and enzymatically.3"5 Som~ patients have li~,er involvement manifested by hepatomegaly and hypoglycemia; a very few patients have muscle weakness and wasting but no clinically apparent liver disease; and many patients have problems related to both liver and muscle. Van Hoof and Hers 3 measured the combined and separate activities of transferase and glucosidase in both liver and muscle, and divided type III disease into six biochemical subgroups, s In this classification, most patients have had very low or absent debranching enzyme activity in both liver and muscle (type I l i A disease), and some have had enzyme deficiency limited to the liver (type IIIB disease). Some patients have had an intermediate level of enzyme activity, and in rare instances a selective loss of only one of the two debranching activities (glucosidase or transferase) has been demonstrated. In a study of 34 patients, Brown and Brown 4 found disease subtypes I l i A and IIIB and one example of possible isolated glucosidase deficiency, but they were unable to identify cases with intermediate levels of enzyme activity. It appears that 75% to 85% of patients with glycogen storage disease type III have generalized deficiency, affecting both liver and muscle (type I l i A disease), and 17% have the enzyme activity absent in the liver but retained in the muscle (type IIIB disease); only rarely do patients have a selective loss of one of the two activities.46 The molecular basis for the clinical and enzymatic variability in glycogen storage disease type III is not understood. We therefore initiated studies of debranching enzyme to elucidate the molecular mechanism and to identify the basis for the subtypes observed enzyrnatically. We previously reported the purification of glycogen debranching enzyme, characterization of antibody, and absence of cross-reactive material in five patients with type I l i a disease. 7 In this study, we extended our investigation of the debranching enzyme to a total of 41 patients with various subtypes, and we now report on the enzyme from a variety of tissues. We also report the first use of immunoblot analysis for the detection of heterozygotes and for prenatal diagnosis. METHODS Reagents. Nitrocellulose paper was purchased from Schleicher & Schuell Inc. (Keene, N.H.); iodine 125labeled protein A was obtained from Amersham Corp. . (Arlington Heights, t11.); and Coomassie brilliant blue, protein-molecular-weight standards, and reagents for polyacrylamide gels were from Bio-Rad Laboratories (Richmond, Calif.). Tissue culture supplies and fetal calf serum were from Gibco Laboratories (Grand Island, N.Y.). Phenylmethyisulfonyl fluoride, ethylenediaminetetraacetate,
and ethyleneglycoltetraacetate were from Sigma Chemical Co. (St. Louis, Mo.). Patients. We studied 41 patients with glycogen storage disease type III whose ages ranged from 4 months to 50 years. Type III disease had been previously diagnosed by demonstrating deficient debranching enzyme activity in liver or muscle biopsy samples, or both, with phosphorylase limit dextrin used as the substrate. 8 This method measures combined oligo-l,4-1,4-glucantransferase and amylo-l,6glucosidase activities. Incorporation of carbon 14-labeled glucose into glycogen was also measured in 15 patients. 9 In 24 cases, debranching enzyme activity was measured in both liver and muscle. By the Van Hoof-Hers enzymatic classification,3 19 patients had disease that involved both liver and muscle (type I l i A ) , four had disease confined to the liver (type IIIB), and three had isolated transferase deficiency in liver and muscle (type IIID disease). Type IIID disease is defined as deficient debranching enzyme activity with phosphorylase limit dextrin but normal or slightly reduced incorporation of [14C] glucose into glycogen. In the remaining 15 patients, debranching enzyme activity was measured only in liver; in 12 of these patients, the clinical findings suggested that muscle involvement was also present; in the other three patients, the clinical data were incomplete. Tissue samples. All human liver and muscle biopsy samples used in our investigation had been stored at - 2 0 ~ C or - 7 0 ~ C (some for >2 years) after the original enzymatic or pathologic diagnosis was made. Control samples were from patients with other metabolic diseases (cystic fibrosis; glycogen storage disease types I, II, VI, and IX; unknown myopathy). Chorionic villi and amniotic cells were obtained from pregnant women whose fetuses were at risk for chromosomal abnormalities. Fresh porcine tissues were obtained from a local abattoir. Cell lines. Four skin fibroblast cultures (GM 02523, G M 03390, GM 00576, GM 00573) from patients with glycogen storage disease type III were obtained from the Institute of Medical Research, Camden, N.J. Six skin fibroblast cultures and eight lymphoblast0id cell lines from patients with type III disease were established in our laboratory after informed consent for skin biopsy and blood drawing was obtained. Lymphocytes were separated from peripheral blood samples and transformed by Epstein-Barr virus to produce permanent cell lines. The fibroblast cultures were maintained in Eagle minimum essential medium with 10% fetal calf serum, and the lymphoblastoid cells were grown in RPMI 1640 culture medium (Gibco) with 10% fetal calf serum. Penicillin, 100 units/ml, and streptomycin, i 00 #g/ ml, were added to the culture media.
Immunoblot analysis Antiserum. The antiserum against purified porcinemuscle debranching enzyme was produced in New Zealand
Volume 116 Number I
G l y c o g e n d e b r a n c h i n g e n z y m e in glycogen s t o r a g e disease
rabbits. This antiserum reacts against porcine debranching enzyme with a single precipitation line and inhibits the debranching enzyme activity] Because our antiserum recognized nonspecific bands in fibroblasts, lymphoblastoid cells, and lymphocytes, the antiserum was affinity-purified with purified porcine enzyme before immunoblot analyses of these cells. Antiserum (0,5 ml) was added to a nitrocellulose disk (1.5 cm in diameter) containing 20 ~zg of purified debranching enzyme and incubated for 1 hour. The filter was washed with phosphate-buffered saline solution, and antibody was eluted with 0.5 sodium chloride, 0.2 mol/L acetic acid, and then neutralized. W e s t e r n blots. Tissue samples were homogenized and cultured cells were sonicated in the presence of 5 volumes of buffer containing 4 mmol/L ethylenediaminetetraacetare, 2 mmol/L ethyleneglycoltetraacetate (pH 7.2), and 0.34 mmol/L phenylmethylsulfonyl fluoride. The homogenate or sonicate was then centrifuged at 10,000g for 15 minutes, an d the supernatant was subjected to immunoblot analysis. Up to 80 ~zg of protein from the supernatant was applied to 7% sodium dodecylsulfate-potyacrylamide gel electrophoresis according to the method of Laemmli) ~ The electrotransfer of protein and the immunologic detection of protein were performed as described elsewhere, 11 with the following modifications: The proteins were electroblotted onto nitocellulose paper by means of transblotter equipment (Bio-Rad Laboratories). The nitrocellulose filter was washed with 5% milk (Carnation nonfat dry milk) in phosphate-buffered saline solution for 1 hour. The antiserum or antibody (1:100 dilution in fresh 5% milk) was then added, and the solution was incubated for 2 hours. The filter was washed for 1 hour with three changes of 1% milk and incubated with 5 X l0 s cpm/ml [1251] protein A for 2 hours. The filter was then washed three times with 1% milk and once with 0.5% Triton X-100 detergent, each time for 15 minutes. The filter was then dried and autoradiographed. Protein was measured by the Bradford method, 12 with bovine serum albumin used as the standard. RESULTS Tissue distribution of debranching enzyme. Immunoblot analysis with anti-porcine-muscle debrancher antiserum showed that this antiserum can detect debranching enzyme in porcine kidney, heart, skeletal muscle, liver, brain, and spleen (data not shown). The major band detected in each of these tissues was 160 kd, the same size as the purified debranching enzyme. The amount of debfancher protein was highest in liver, heart, and skeletal muscle, and low in spleen, kidney, and brain. The antibody als0 detected debrancher protein in cultured chorionic villi cells, lymphocytes, lymphoblastoid cells, skin fibroblasts, and cultured amniotic fluid cells (Fig. 1). The amount of debrancher protein in these tissues was small in comparison with that
]
2
3
4
5
6
7
97
8
tOO
16 -93
-66
,~
-45
Fig. t. Distribution of debranching enzyme m human tissues obtained by immunoblot analysis of protein from various human tissues probed with affinity-purified antibody against muscle debranching enzyme, Lane I, Erythrocytes; lane 2, cultured chorionic villi ceils; lane 3, lymphoeytes; lane 4, lymphoblastoid cells; lane 5, muscle tissue; lane 6, cultured fibroblasts; lane 7, cultured amniotic cells; and lane 8, liver tissue. In lanes I to 4, 6, and 7, each well contained 80 gg of protein. Lanes 5 (muscle) and 8 (liver) each contained 20/zg of protein. Numbers on right are molecular masses, measured in kilodaltons, of marker proteins.
in liver or muscle. As estimated from densitometry scanning and the amount of protein applied to the immunoblot, the amount of debrancher protein in these tissues was less than 10% of the amount that was present in muscle or liver. Debrancher protein in patients with glycogen storage disease type IIL Forty-eight samples from 41 patients with glycogen storage disease type III were subjected to immunoblot analysis. The analyses were performed on coded samples without prior knowledge of the diagnoses. The samples were from 16 livers, 10 muscles, 3 lymphocyte preparations, 10 skin fibroblast cultures, and 8 lymphoblastold cell lines. In 7 patients, more than one tissue was analyzed: liver and muscle, 4 patients (3 with type I I I A disease and 1 with type IIID disease); fibroblasts and lymphoblastoid cells, 2 patients (type IIIB disease); lymphocytes and fibroblasts, 1 patient (type IIIB disease). Representative samples are shown in Figs, 2 and 3. Deficient debranching enzyme was demonstrated, not only in liver, muscle, or
98
Ding et al.
The Journal o f Pediatrics January 1990
MUSCLE I
2
3
4
LIVER 5
6
7
8
LIVER 9
I0
It
MUSCLE
12 15 14 15 16 17
%:
..... ,,
Fig. 2. Immunoblot analysis of debrancher protein from liver and muscle tissue probed with antiserum against porcine debranching enzyme. Each well contained 40 ~g of protein. Tissues contained in lanes 2, 3, 4, 6, 7, 13, 14, and 17 were from patients with glycogen storage disease type II[. Lane 6 contains tissue from patient with isolated transferase deficiency (type IIID disease), lane 14 from patient with type IIIB disease, and remainder from patient with type IIIA disease. Lane 7 (liver) and lane 17 (muscle) contain tissues from same patient with type IIIA disease. Remaining lanes contain control samples from patients with other metabolic diseases. Arrow indicates position of debranching enzyme.
F I BRQBLASTS 2
~
a
5
6
7
8
9
LYMPHOCYTES t0
~
12
15
~4
15
16
17
IB
19
20
Fig. 3. Immunoblot analysis of debrancher protein from skin fibroblasts and lymphocytes probed with affinity-purified antibody against muscle debranching enzyme. Each well contained 40 #g protein from skin fibroblasts ,lanes 1 to 6 and 8 to 14) or lymphocytes (lanes 15 to 20). Lane 7 contained 10 #g of protein from human muscle as control. Lanes 1, 4, 5, 6, 12, 14, and 18 contained tissue from patients with type IIIA disease, and lanes 9 and 15 from type IIIB disease. Lanes 10 and 11 and lanes 19 and 20 were from parents of patients 12 and 18, respectively. Lanes 2, 3, 8, 13, 16, and 17 are control samples from patients with other metabolic diseases.
both, but also in skin fibroblasts, lymphocytes, and lymphoblastoid cells. In three patients who had nearly normal amounts of CRM, isolated transferase deficiency (type IIID) was diagnosed enzymatically. The patient with type IIID disease, whose findings are shown in Fig. 2, also had C R M present in a muscle sample. In four patients with type IIIB disease, C R M was absent in all four skin fibroblast cultures, in both of the lymphoblastoid cell lines, and in the lymphocyte culture studied. No liver or muscle samples from the four patients with type IIIB disease were available.
The patients with type I I I A disease included 19 with an enzymatic diagnosis made on the basis of both liver and muscle samples and 12 whose diagnosis was made only on the basis of liver samples but who had clinical findings compatible with muscle involvement. In these patients, immunoblot analysis was performed on skin fibroblasts, lymphoblastoid cells, and lymphocytes in 18, and on liver or muscle samples, or both, in the remaining 13 patients. Regardless of the tissue source studied, all patients with type I I I A disease had either absent or a greatly reduced amount
Volume 116 Number l
Glycogen debranching enzyme in glycogen storage disease
of CRM representing debrancher protein. In two patients who had residual CRM in muscle, the debranching enzyme band was no more than 5% of the control by densitometry scanning. We did not detect debranching enzyme of an abnormal size in any patient. For an additional three patients, insufficient data were available to determine muscle status; enzyme activity was deficient in liver, and CRM was absent.
Heterozygote detection and prenatal diagnosis of glycogen storage disease type IIl. Skin fibroblasts and lymphocytes from two families with type I I I A disease were subjected to immunoblot analysis. No debranching enzyme band was detected in the probands (Fig. 3). The parents of the patients had an intermediate level of debranching enzyme protein in cell extracts, consistent with their presumed hcterozygote state (Fig. 3). Prenatal diagnosis was performed in one of the families. In early childhood the proband had had hypoglycemia, hepatomegaly, and failure to thrive. Cultured amniocytes from the pregnancy at risk demonstrated a debranching enzyme band of the same size and intensity as a control amniocyte culture, thus predicting an unaffected fetus (data not shown). This result was confirmed postnatally by immunoblot analysis of the lymphoblastoid cells established after birth. The baby at 1 year of age had no clinical or laboratory evidence of glycogen storage disease type III. DISCUSSION Most patients with glycogen storage disease type III manifest the disease in several organ systems. Symptoms of myopathy and of liver involvement are most frequent, but neural involvement, renal tubular acidosis, splenomegaly, and cardiac enlargement have been described. Although glycogen with short outer branches accumulates in erythrocytes of patients with type III disease, the absence of debranching enzyme CRM in normal erythrocytes suggests that the amount of debranching enzyme in these cells is lower than our method can detect. The presence of debranching enzyme in a variety of tissues is consistent with multiorgan system involvement in type III disease. Regardless of whether enzyme deficiency or clinical disease was identified in both liver and muscle (type I I I A disease) or was confined to liver alone (type IIIB disease), our data demonstrated the skin fibroblasts, lymphocytes, and lymphoblastoid cells had absent CRM. This finding suggests that the regulation of dehranching enzyme expression is similar in liver, fibroblasts, lymphocytes, and lymphoblastoid cells but is different in muscle. The molecular mechanism for the control of tissue-specific expression of debranchifig enzyme is unknown. Regulation is unlikely to be due to the presence of multiple genes that encode isoenzymes, because previous studies of the purified protein provide no evidence for subunits or isoenzymes of debranching enzyme, and because
99
our data show that the debranching enzyme is 160 kd in all tissues examined. Isolated transferase deficiency (type IIID disease) was diagnosed by measuring deficient debranching enzyme activity with phosphorylase limit dextrin but normal or only slightly reduced incorporation of [14C]glucose into glycogen. Although not all patients studied had enzyme activity measured by both methods, we identified three patients who had nearly normal amounts o] CRM, all of whom had isolated transferase deficiency (type IIID disease). The presence of CRM indicates that the principal antigenic determinant regions of the protein are not located near the catalytic center responsible for transferase activity. Thus our demonstration that patients with type IIID disease have CRM of 160 kd strongly supports previous studies which indicated that debranching enzyme contains two catalytic activities at separate sites on a single polypeptide chain. I, 2 At the molecular level, the defects causing absent CRM may be heterogeneous. The mutation may affect the promotor region or splicing unction. The mutation may create a nucleotide substitution that results in an incomplete protein or an unstable protein. An additional possibility is that the absence of CRM may result from a major alteration of the debranching enzyme gene. In any case, the fact that C R M is absent in most patients wtih glycogen storage disease type II1 has a practical implication. Because enzyme activity is relatively low and difficult to measure in readily accessible tissues such as skin fibroblasts, cultured amniotic fluid cells, or leukocytes,6, 13, 14 Western blot analysis offers an alternative method for prenatal and postnatal diagnosis of type III disease, provided that the proband has no CRM for debranching enzyme. If the CRM status of the proband is not known, a combination immunobtot and biochemical analysis 14-16would be most helpful in providing the diagnosis. Although type III disease can be diagnosed in readily accessible tissues, a positive test result will not indicate whether the enzyme deficiency is generalized or confined to the liver. For the initial diagnosis, we recommend that fibroblasts be sampled. At the present time we do not know whether a normal serum creatine kinase level excludes muscle involvement. Therefore the differentiation between type I I I A and IIIB disease will also require analysis of material obtained by muscle biopsy. We are grateful to the Genetic Division of New York University Medical Center for tissue contribution.
REFERENCES 1. Bates EJ, Heaton GM, Taylor C, Kernohan JC, Cohen P. Debranching enzyme from rabbit skeletal muscle: evidence for the location of two active centres on a single polypeptide chain. FEBS Lett 1975;58:181-5. 2. Gillard BK, Nelson TE. Amylo-l,6-glucosidase/4-a-glucan-
10 0
3. 4.
5.
6.
7.
8.
9.
Ding et al.
The Journal of Pediatrics January 1990
otransferase: use of reversible substrate model inhibitors to study the binding and active sites of rabbit muscle debranehing enzyme. Biochemistry 1977;16:3978-87. Van Hoof F, Hers HG. The subgroups of type llI glycogenosis. Biochemistry 1967;2:265-70. Brown BI, Brown DH. Glycogen storage diseases: types I, III, IV, V, VII and unclassified glycogenoses. In: Dickens F, Randie P J, Whelan W J, eds. Carbohydrate metabolism and its disorders; vol 2. New York: Academic Press, 1968;123-60. Howell RR, Williams JC. The glycogen storage diseases. In: Stanbury JB, Wyngaarden JB, Fredrickson DS, Goldstein JL, Brown MS, eds. The metabolic basis of inherited disease. 5th ed. New York: McGraw-Hill, 1983;141-66. Brown BI, Brown DH. Definitive assays for glycogen debranching enzyme in human fibroblasts. In: Schotland DL, ed. Disorders of the motor unit. New York: Wiley, 1982;667-73. Chen YT, He JK, Ding JH, Brown BI. Glycogen debranching enzyme: purification, antibody characterization, and immunoblot analyses of type III glycogen storage disease. Am J Hum Genet 1987;41:1002-15. Brown DH, Brown BI. Enzymes of glycogen debranching: amylo-l,6-glucosidase (I) and oligo-l,4-glucantransferase (II). Methods Enzymol 1966;8:515-24. Hers HG, Verhue W, Van Hoof F. The determination of amylo-l,6-glucosidase. Eur J Biochem 1967;2:257-64.
10. Laemmli UK. Cleavage ofstructural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227: 680-5. 11. Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedures and' some applications. Proc Natl Acad Sci USA 1979;76:4350-4. 12. Bradford M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Chem 1976;72:248-54. 13. Gutman A, Barash V, Schramm H, et al. Incorporation of [14C]glucose into ~-1,4 bonds of glycogen by leukocytes and fibroblasts of patients with type III glycogen storage disease. Pediatr Res 1985;19:28-32. 14. Van Diggelen OP, Janse HC, Smit GPA. Debranching enzyme in fibroblasts, amniotic fluid cells and chorionic villi: pre- and postnatal diagnosis of glycogenosis type III. Clin Chim Acta 1985;149:129-34. 15. Brown BI, Brown DH. Definitive assays for glycogen debranching enzyme in human fibroblasts. In: Schotland DL, ed. Disorders of the motor unit. New York: Wiley, 1982:667-73. 16. Maire I, Mathieu M. Possible prenatal diagnosis of type III glycogenosis. J Inher Metab Dis 1986;9:89-91.
FELLOWSHIPS Available fellowships in pediatric subspecialties and those for general academic pediatric training are listed once a year, in May, in THE JOURNAL OF PEDIATRICS. Each October, forms for listing such fellowships are sent to the Chairman of the Department of Pediatrics at most major hospitals in the United States and Canada. Should you desire to list fellowships, a separate application must be made each year for each position. All applications must be returned to The C.V. Mosby Company by February 15 of the listing year to ensure publication. Additional forms will be supplied on request from the Journal Editing Department, The C.V. Mosby Company, 11830 Westline Industrial Drive, St. Louis, M O 63146-3318/314-872-8370.